• Journal of KIISE:Software and Applications
• /
• v.26 no.10
• /
• pp.1204-1218
• /
• 1999

문서영상 검색을 위한 디지털도서관의 대부분은 논문제목과/또는 논문요약으로부터 만들어진 색인에 근거한 제한적인 검색기능을 제공하고 있다. 본 논문에서는 영문 문서영상전체에 대한 검색을 위한 단어 영상 형태 특징기반의 단어검색시스템을 제안한다. 본 논문에서는 검색의 효율성과 정확도를 높이기 위해 1) 기존의 단어검색시스템에서 사용된 특징들을 조합하여 사용하며, 2) 특징의 개수 및 위치뿐만 아니라 특징들의 순서를 포함하여 매칭하는 방법을 사용하며, 3) 특징비교에 의해 검색결과를 얻은 후에 여과목적으로 문자인식을 부분적으로 적용하는 2단계의 검색방법을 사용한다. 제안된 시스템의 동작은 다음과 같다. 문서 영상이 주어지면, 문서 영상 구조가 분석되고 단어 영역들의 조합으로 분할된다. 단어 영상의 특징들이 추출되어 저장된다. 사용자의 텍스트 질의가 주어지면 이에 대응되는 단어 영상이 만들어지며 이로부터 영상특징이 추출된다. 이 참조 특징과 저장된 특징들과 비교하여 유사한 단어를 검색하게 된다. 제안된 시스템은 IBM-PC를 이용한 웹 환경에서 구축되었으며, 영문 문서영상을 이용하여 실험이 수행되었다. 실험결과는 본 논문에서 제안하는 방법들의 유효성을 보여주고 있다. Abstract Most existing digital libraries for document image retrieval provide a limited retrieval service due to their indexing from document titles and/or the content of document abstracts. This paper proposes a word spotting system for full English document image retrieval based on word image shape features. In order to improve not only the efficiency but also the precision of a retrieval system, we develop the system by 1) using a combination of the holistic features which have been used in the existing word spotting systems, 2) performing image matching by comparing the order of features in a word in addition to the number of features and their positions, and 3) adopting 2 stage retrieval strategies by obtaining retrieval results by image feature matching and applying OCR(Optical Charater Recognition) partly to the results for filtering purpose. The proposed system operates as follows: given a document image, its structure is analyzed and is segmented into a set of word regions. Then, word shape features are extracted and stored. Given a user's query with text, features are extracted after its corresponding word image is generated. This reference model is compared with the stored features to find out similar words. The proposed system is implemented with IBM-PC in a web environment and its experiments are performed with English document images. Experimental results show the effectiveness of the proposed methods.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.10 no.4
• /
• pp.351-356
• /
• 2000

The important thing of Information Retrieval System is to satisfy is to satisfy the user's requriement in searching Information Retrieval system ranks documents by weights in document, then Retrieved document context does not consist with given query. This paper proposes a new method of document retrieval based on extended fuzzy concept networks. there are four of fuzzy relationships between concept; fuzzy positive combination, fuzzy negative combination, fuzzy generalization, and fuzzy specilalization. After modeling an extended fuzzy concept network by relation matrix and relevance matrix, we measured similarties.

• Annual Conference on Human and Language Technology
• /
• 2022.10a
• /
• pp.135-140
• /
• 2022

최근 Document Retrieval을 비롯한 대부분의 자연어처리 분야에서는 BERT와 같이 self-attention을 기반으로 한 사전훈련 모델을 활용하여 SOTA(state-of-the-art)를 이루고 있다. 그러나 self-attention 메커니즘은 입력 텍스트 길이의 제곱에 비례하여 계산 복잡도가 증가하기 때문에, 해당 모델들은 선천적으로 입력 텍스트의 길이가 제한되는 한계점을 지닌다. Document Retrieval 분야에서는, 문서를 특정 토큰 길이 단위의 문단으로 나누어 각 문단의 유사 점수 또는 표현 벡터를 추출한 후 집계함으로서 길이 제한 문제를 해결하는 방법론이 하나의 주류를 이루고 있다. 그러나 논문, 특허와 같이 섹션 형식(초록, 결론 등)을 갖는 문서의 경우, 섹션 유형에 따라 고유한 정보 특성을 지닌다. 따라서 문서를 단순히 특정 길이의 문단으로 나누어 학습하는 PARADE와 같은 기존 방법론은 각 섹션이 지닌 특성을 반영하지 못한다는 한계점을 지닌다. 본 논문에서는 섹션 유형에 대한 정보를 포함하는 문단 표현을 학습한 후, 트랜스포머 인코더를 사용하여 집계함으로서, 결과적으로 섹션의 특징과 상호 정보를 학습할 수 있도록 하는 SERADE 모델을 제안하고자 한다. 실험 결과, PARADE-Transformer 모델과 비교하여 평균 3.8%의 성능 향상을 기록하였다.

• Journal of Information Science Theory and Practice
• /
• v.12 no.3
• /
• pp.36-48
• /
• 2024

The information retrieval (IR) process often encounters a challenge known as query-document vocabulary mismatch, where user queries do not align with document content, impacting search effectiveness. Automatic query expansion (AQE) techniques aim to mitigate this issue by augmenting user queries with related terms or synonyms. Word embedding, particularly Word2Vec, has gained prominence for AQE due to its ability to represent words as real-number vectors. However, AQE methods typically expand individual query terms, potentially leading to query drift if not carefully selected. To address this, researchers propose utilizing median vectors derived from deep median networks to capture query similarity comprehensively. Integrating median vectors into candidate term generation and combining them with the BM25 probabilistic model and two IR strategies (EQE1 and V2Q) yields promising results, outperforming baseline methods in experimental settings.

• International Journal of Computer Science & Network Security
• /
• v.24 no.7
• /
• pp.31-44
• /
• 2024

The inverted index is combination of the keywords and posting lists associated for indexing of document. In modern age excessive use of technology has increased data volume at a very high rate. Big data is great concern of researchers. An efficient Document indexing in big data has become a major challenge for researchers. All organizations and web engines have limited number of resources such as space and storage which is very crucial in term of data management of information retrieval system. Information retrieval system need to very efficient. Inverted indexing technique is introduced in this research to minimize the delay in retrieval of data in information retrieval system. Inverted index is illustrated and then its issues are discussed and resolve by implementing the scalable inverted index. Then existing algorithm of inverted compared with the naïve inverted index. The Interval list of inverted indexes stores on primary storage except of auxiliary memory. In this research an efficient architecture of information retrieval system is proposed particularly for unstructured data which don't have a predefined structure format and data volume.

• Journal of the Korean Society for Library and Information Science
• /
• v.17
• /
• pp.49-83
• /
• 1989

A Full text retrieval system is a natural language document retrieval system in which the full text of all documents in a collection is stored on a computer so that every word in every sentence of every document can be located by the machine. This kind of IR System is recently becoming rapidly available online in the field of legal, newspaper, journal and reference book indexing. Increased research interest has been in this field. In this paper, research on full text databases and retrieval systems are reviewed, directions for research in this field are speculated, questions in the field that need answering are considered, and variables affecting online full text retrieval and various role that variables play in a research study are described. Two obvious research questions in full text retrieval have been how full text retrieval performs and how to improve the retrieval performance of full text databases. Research to improve the retrieval performance has been incorporated with ranking or weighting algorithms based on word occurrences, combined menu-driven and query-driven systems, and improvement of computer architectures and record structure for databases. Recent increase in the number of full text databases with various sizes, forms and subject matters, and recent development in computer architecture artificial intelligence, and videodisc technology promise new direction of its research and scholarly growth. Studies on the interrelationship between every elements of the full text retrieval situation and the relationship between each elements and retrieval performance may give a professional view in theory and practice of full text retrieval.

• Journal of KIISE:Computing Practices and Letters
• /
• v.13 no.6
• /
• pp.407-411
• /
• 2007

Over the years, great attention has been paid to the question of exploiting inherent semantic of HTML in the area of web document retrieval. Although HTML is mainly presentation oriented, HTML tags implicitly contain useful semantics that can be catch meaning of text. Focusing on this idea. in this paper we define 'markup pattern' and try to improve performance of web document retrieval using markup patterns. Markup pattern is a mirror of intends of web document publisher and an internal semantic of text on web document. To discover the markup pattern and exploit it, we suggest a new scheme for extracting concepts and weighting documents. For evaluation task, we select two domains-BBC and CNN web sites, and use their search engines to gather domain documents. We re-weight and re-score documents using proposed scheme, and show the performance improvement in the two domains.

• Journal of KIISE:Software and Applications
• /
• v.28 no.10
• /
• pp.773-784
• /
• 2001

Since documents on the Web are naturally partitioned into may document database, the efficient information retrieval process requires identifying the document database that are most likely to provide relevant documents to the query and then querying the identified document database. We propose a neural net agent approach to such an efficient information retrieval. First, we present a neural net agent that learns about underlying document database using the relevance feedbacks obtained from many retrieval experiences. For a given query, the neural net agent, which is sufficiently trained on the basis of the BPN learning mechanism, discovers the document database associated with the relevant documents and retrieves those documents effectively. In the experiment, we introduce a neural net agent based information retrieval system and evaluate its performance by comparing experimental results to those of the conventional well-known approaches.

• Journal of Multimedia Information System
• /
• v.4 no.3
• /
• pp.99-106
• /
• 2017

This study does a qualitative and quantitative analysis of how music by humming is being used by music educators in the classroom. Music by humming is part division of music information retrieval. In order to define what a music information retrieval system is first I need to define what it is. Berger and Lafferty (1999) define information retrieval as "someone doing a query to a retrieval system, a user begins with an information need. This need is an ideal document- perfect fit for the user, but almost certainly not present in the retrieval system's collection of documents. From this ideal document, the user selects a group of identifying terms. In the context of traditional IR, one could view this group of terms as akin to expanded query." Music Information Retrieval has its background in information systems, data mining, intelligent systems, library science, music history and music theory. Three rounds of surveys using question pro where completed. The study found that there were variances in knowledge, training and level of awareness of query by humming, music information retrieval systems. Those variance relationships where based on music specialty, level that they teach, and age of the respondents.

• Journal of the Korean Society for information Management
• /
• v.19 no.3
• /
• pp.91-110
• /
• 2002

Recently, to overcome the limit of keyword based retrieval system, the study based structured document has been studied. But it is hard for structured retrieval system to adapt a real service, in this paper, we propose a method of retrieval mechanism for the fulltext of XML documents. We explain DTD of XML based report, extracting navigation information and planing to adapt the retrieval system for article retrieval. Using the fulttext retrieval scheme, suggested system can be an alternative plan of professional structured based retrieval system.